A combination of synthetic chemical technologies and certain computer-related technologies has led to the development of an important analytical tool in the field of molecular biology commonly referred to as the "gene chip." Gene chips are high-density arrays of biopolymers, such as oligonucleotides or complementary deoxyribonucleic acid ("cDNA") molecules bound to a chemically prepared substrate such as silicon, glass, or plastic. In the case of oligonucleotide-containing molecular arrays, each cell, or element, within an array is prepared to contain a single oligonucleotide species, and the oligonucleotide species in a given cell may differ from the oligonucleotide species in the remaining cells of the high-density array. Gene chips may be used in DNA hybridization experiments in which radioactively, fluorescently, or chemiluminescently labeled deoxyribonucleic ("DNA") or ribonucleic acid ("RNA") molecules are applied to the surface of the gene chip and are bound, via Watson-Crick base pair interactions, to specific oligonucleotides bound to the gene chip. The gene chip can then be analyzed by radiometric or optical methods to determine to which specific cells of the gene chip the labeled DNA or RNA molecules are bound. Thus, in a single experiment, a DNA or RNA molecule can be screened for binding to tens or hundreds of thousands of different oligonucleotides.
Hybridization experiments can be used to identify particular gene transcripts in messenger RNA ("mRNA") preparations, to identify the presence of genes or regulatory sequences in cDNA preparations, or to sequence DNA and RNA molecules. Particularly in the latter application, the effectiveness of employing gene chips depends of the precision with which specific oligonucleotides can be synthesized within discrete cells of the gene chip. As with any chemical synthetic process, various factors may cause the yields of specific steps in the synthesis of oligonucleotides to be less than 100%, leading to unintended and unwanted intermediate species.
During an oligonucleotide initiation on lengthening step in the synthesis of oligonucleotides on the surface of a gene chip, reactive deoxynucleoside phosphoramidites are successively applied, in concentrations exceeding the concentrations of target hydroxyl groups of the substrate or growing oligonucleotide polymers, to specific cells of the high-density array, where they chemically bond to the target hydroxyl groups. Then, unreacted deoxynucleoside phosphoramidites from multiple cells of the high-density array are washed away, oxidation of the phosphite bonds joining the newly added deoxynucleosides to the growing oligonucleotide polymers to form phosphate bonds is carried out, and unreacted hydroxyl groups of the substrate or growing oligonucleotide polymers are chemically capped to prevent them from reacting with subsequently applied deoxynucleoside phosphoramidites. The chemical capping agents currently employed have rather low efficiencies for capping substrate hydroxyl groups, and better, but not extremely high, efficiencies for capping unreacted growing oligonucleotide polymers. As a result, each element of a molecular array may end up containing significant amounts of undesirable polymers that are unintentionally synthesized along with a desired polymer.
Molecular arrays are prepared using highly automated methods in which a series of droplets, each containing one particular type of reactive deoxynucleoside phosphoramidite, is sequentially applied to each element by a mechanical device, such as an inkjet print head. Unfortunately, the precision at which successive droplets can be applied to an element is insufficient to guarantee that each successive droplet will be confined within the boundaries of the element, or that the entire element will be covered by any particular droplet. Misregistration of successively applied droplets, like the use of inefficient capping agents, may lead to significant amounts of undesirable polymers that are unintentionally synthesized along with a desired polymer within each element, and may, in addition, lead to synthesis of unwanted polymers in regions of the surface of the molecular array substrate adjacent to each element.
The presence of undesirable polymers may lead to less specific binding of radioactively, fluorescently, or chemiluminescently labeled DNA or RNA molecules to molecular array elements, in turn leading to a significant decrease in the signal-to-noise ratio in the analysis of the molecular array and leading to spurious results. In order to automatically scan molecular arrays for the presence of radioactively, fluorescently, or chemiluminescently labeled DNA or RNA molecules, it is most desirable for the surface of the elements to be uniformly covered with desired substrate-bound polymers, and for each element to have a sharply defined edge. The inter-element surface of the molecular array substrate should have little or no contaminants that can bind DNA or RNA, including substrate-bound polymers inadvertently synthesized along with the polymers synthesized within the elements. Otherwise, after exposure of the molecular array to labeled sample molecules, fuzzy, indistinct areas of the molecular array substrate will contain labeled DNA or RNA, making it difficult for feature extraction software to select an area corresponding to an element over which to average signal intensity. Poorly averaged signal intensity may significantly lower confidence in resulting measurements, and may even produce incorrect results. Manufacturers of molecular arrays, and experimentalists and diagnosticians using molecular arrays, have therefore recognized a need for techniques for more specifically synthesizing desired polymers within molecular array elements and for uniformly covering the surface of molecular array elements with desired polymers, so that each element has a clearly defined edge and so that the inter-element regions of the molecular array substrate have as small amounts of undesirable sample-molecule-binding substances as possible.